On Wed, 16 Sep 2015, Dan Carpenter wrote:
> The cinfo struct has a hole after the last struct member so we need to
> zero it out. Otherwise we don't disclose some uninitialized stack data.
I think the "don't" wasn't intended in the second sentence?
julia
>
> Signed-off-by: Dan Carpenter
The cinfo struct has a hole after the last struct member so we need to
zero it out. Otherwise we don't disclose some uninitialized stack data.
Signed-off-by: Dan Carpenter
diff --git a/drivers/staging/rdma/hfi1/file_ops.c
b/drivers/staging/rdma/hfi1/file_ops.c
index
mutex_trylock() returns zero on failure, not EBUSY.
Signed-off-by: Dan Carpenter
diff --git a/drivers/staging/rdma/hfi1/chip.c b/drivers/staging/rdma/hfi1/chip.c
index 654eafe..aa58e59 100644
--- a/drivers/staging/rdma/hfi1/chip.c
+++ b/drivers/staging/rdma/hfi1/chip.c
Hello Bryan O'Sullivan,
The patch 6522108f19a9: "IB/ipath: infiniband verbs support" from Mar
29, 2006, leads to the following static checker warning:
drivers/staging/rdma/ipath/ipath_verbs.c:2289 show_hca()
warn: bool is not less than zero.
The cinfo struct has a hole after the last struct member so we need to
zero it out. Otherwise we disclose some uninitialized stack data.
Signed-off-by: Dan Carpenter
---
v2: typo in changelog
diff --git a/drivers/staging/rdma/hfi1/file_ops.c
On Wed, Sep 16, 2015 at 08:25:00AM +0200, Julia Lawall wrote:
> On Wed, 16 Sep 2015, Dan Carpenter wrote:
>
> > The cinfo struct has a hole after the last struct member so we need to
> > zero it out. Otherwise we don't disclose some uninitialized stack data.
>
> I think the "don't" wasn't
__get_txreq() returns an ERR_PTR() but this checks for NULL so it would
oops on failure.
Signed-off-by: Dan Carpenter
diff --git a/drivers/staging/rdma/hfi1/verbs.c
b/drivers/staging/rdma/hfi1/verbs.c
index 53ac214..41bb59e 100644
---
Hello Mike Marciniszyn,
The patch 7724105686e7: "IB/hfi1: add driver files" from Jul 30,
2015, leads to the following static checker warning:
drivers/staging/rdma/hfi1/user_sdma.c:1349 set_txreq_header_ahg()
warn: mask and shift to zero
drivers/staging/rdma/hfi1/user_sdma.c
Hello Mike Marciniszyn,
The patch 7724105686e7: "IB/hfi1: add driver files" from Jul 30,
2015, leads to the following static checker warning:
drivers/staging/rdma/hfi1/rc.c:2399 hfi1_rc_hdrerr()
warn: right shift assign to zero
drivers/staging/rdma/hfi1/rc.c
2376 void
From: Amir Vadai
Adding a macro for RoCEv2 UDP destination port.
Signed-off-by: Amir Vadai
Signed-off-by: Kamal Heib
---
include/rdma/ib_verbs.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/include/rdma/ib_verbs.h
Manage and allocate pool of objects with given limit on number of
elements. Gets parameters from rxe_type_info. Pool elements are
allocated out of a slab cache. Objects that are using this facility
are: PD, QP, SRQ, CQ, MR, FMR, MW, etc.
Signed-off-by: Kamal Heib
Functions to manipulate SRQ.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_srq.c | 195 ++
1 file changed, 195 insertions(+)
A 'task' is a short function that returns 0 as long as it needs to be
called again. rxe tasks are based on the kernel's tasklet infrastructure.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
Kconfig and Makefiles for RXE driver
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/Kconfig | 2 ++
drivers/staging/Makefile | 1 +
drivers/staging/rxe/Kconfig | 23
Functions to manipulate Address Vector.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_av.c | 87
1 file changed, 87
Useful information about work request opcodes and pkt opcodes in table
form.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_opcode.c | 961
Functions to be called by the networking layer.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe.h | 70 +++
1 file changed, 70
Functions to manipulate QP objects.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
---
drivers/staging/rxe/rxe_qp.c | 835 +++
1 file changed, 835 insertions(+)
create mode 100644
From: Amir Vadai
Objects used by the userspace to post work requests.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
include/uapi/rdma/Kbuild | 1 +
Doug and list Hi,
This patchset introduces Soft RoCE driver.
Some background on the driver: The original Soft-RoCE driver was implemented by
Bob Pearson from SFW. Bob started the submission process [3], but his work was
abandoned after v2.
Mellanox decided to pick it up and continue the
From: Amir Vadai
Things todo in order to get out of staging subtree.
Signed-off-by: Amir Vadai
Signed-off-by: Kamal Heib
---
drivers/staging/rxe/TODO | 15 +++
1 file changed, 15 insertions(+)
create mode 100644
QP request logic.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
---
drivers/staging/rxe/rxe_req.c | 679 ++
1 file changed, 679 insertions(+)
create mode 100644 drivers/staging/rxe/rxe_req.c
diff --git
On Wed, Sep 16, 2015 at 04:42:36PM +0300, Kamal Heib wrote:
> Intorduce Add SEND_LAST_INV and SEND_ONLY_INV opcodes in ib_pack.h to be
> used by RXE for RC.
Why does RXE need new public opcodes?
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to
Linux netdev related code
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
---
drivers/staging/rxe/rxe_net.c | 705 ++
drivers/staging/rxe/rxe_net.h | 72 +
2 files changed, 777 insertions(+)
create mode
MR objects handling.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
---
drivers/staging/rxe/rxe_mr.c | 764 +++
1 file changed, 764 insertions(+)
create mode 100644 drivers/staging/rxe/rxe_mr.c
diff --git
mmap routines
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_mmap.c | 173 +
1 file changed, 173 insertions(+)
create mode 100644
Add maintainer for rxe driver
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
MAINTAINERS | 9 +
1 file changed, 9 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index
Intorduce Add SEND_LAST_INV and SEND_ONLY_INV opcodes in ib_pack.h to be
used by RXE for RC.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
include/rdma/ib_pack.h | 4
1 file changed, 4
Default/initial rxe device parameter settings.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_param.h | 177
1 file changed, 177
A simple circular buffer that can optionally be shared between user
space and the kernel and can be resized.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_queue.c | 217
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_loc.h | 291 ++
1 file changed, 291 insertions(+)
create mode 100644
header bit mask definitions and header lengths declaration of the
rxe_opcode_info struct and rxe_wr_opcode_info struct.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
Functions to manipulate CQ.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_cq.c | 165 +++
1 file changed, 165 insertions(+)
On 8/7/2015 4:00 PM, Matan Barak wrote:
This purpose of this series is to add usage of the GID cache to
the CMA and IB stack. Instead of passing Ethernet L2 attributes
via QP attributes, we could just use the GID cache that's already
points to a ndev and thus to all required L2 attributes.
On 9/16/2015 4:43 PM, Kamal Heib wrote:
From: Amir Vadai
Things todo in order to get out of staging subtree.
Signed-off-by: Amir Vadai
Signed-off-by: Kamal Heib
---
drivers/staging/rxe/TODO | 15 +++
1 file changed,
Handling of Work Completions.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
---
drivers/staging/rxe/rxe_comp.c | 728 +
1 file changed, 728 insertions(+)
create mode 100644 drivers/staging/rxe/rxe_comp.c
Handles receiving new packets which are sent to either request or
response processing.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_recv.c | 371
Multicast groups handling.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_mcast.c | 190
1 file changed, 190 insertions(+)
QP response logic.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
---
drivers/staging/rxe/rxe_resp.c | 1368
1 file changed, 1368 insertions(+)
create mode 100644 drivers/staging/rxe/rxe_resp.c
diff --git
sysfs interface for ib_rxe
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_sysfs.c | 168
1 file changed, 168 insertions(+)
Dummy DMA processing for RXE device.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_dma.c | 166 ++
1 file changed, 166
Compute ICRC for UDP/IP/BTH headers
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
---
drivers/staging/rxe/rxe_icrc.c | 96 ++
1 file changed, 96 insertions(+)
create mode 100644
Add declarations for data structures used to hold per opcode
and per work request opcode tables.
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_hdr.h | 950
Module main for ib_rxe
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe.c | 434 ++
drivers/staging/rxe/rxe.h | 2 +
2 files
rxe interface to rdma/core
Signed-off-by: Kamal Heib
Signed-off-by: Amir Vadai
Signed-off-by: Haggai Eran
---
drivers/staging/rxe/rxe_verbs.c | 1429 +++
drivers/staging/rxe/rxe_verbs.h | 496
On 9/16/2015 4:42 PM, Kamal Heib wrote:
Doug and list Hi,
This patchset introduces Soft RoCE driver.
Thanks guys,
Should probably mention that iser initiator was tested over this driver
and works pretty cool! (user-space TGT iser target was tested with librxe).
Cheers,
Sagi.
--
To
On Wed, 16 Sep 2015, Doug Ledford wrote:
> > Abusing it for send-side is probably the wrong
> > direction overall.
>
> I wouldn't "abuse" it for such, I would suggest adding a proper notion
> of send-only registrations.
That is really not necessary for IP traffic. There is no need to track
these
Add a device capability field csum_cap to denote IPv4 checksum offload
support. Devices should configure this field if they support
insertion/verification of IPv4, TCP and UDP checksums on outgoing/incoming
IPv4 packets according link layer and QP types.
Flags IBV_SEND_IP_CSUM and
On 09/15/2015 07:53 PM, Christoph Lameter wrote:
> On Tue, 15 Sep 2015, Jason Gunthorpe wrote:
>
>> The mcast list in the core is soley for listing subscriptions for
>> inbound - ie receive. Abusing it for send-side is probably the wrong
>> direction overall.
>
> Ok then a simple approach would
The checksum offload capability reporting is enabled based on extended verbs.
The capability field has sub-fields for every link layer, and depends on device
cap, each link layer will support specific QP types. These will be reported to
user space.
I'm new to uverbs extensions and looking forward
Signed-off-by: Bodong Wang
---
drivers/infiniband/hw/mlx4/main.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/infiniband/hw/mlx4/main.c
b/drivers/infiniband/hw/mlx4/main.c
index 8be6db8..a70ca6a 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++
New field csum_cap is added to respective uverbs counterpart according
to ib_verbs.
Signed-off-by: Bodong Wang
---
drivers/infiniband/core/uverbs_cmd.c | 7 +++
include/uapi/rdma/ib_user_verbs.h| 6 ++
2 files changed, 13 insertions(+)
diff --git
On Wed, 16 Sep 2015, Doug Ledford wrote:
> On 09/15/2015 07:53 PM, Christoph Lameter wrote:
> > On Tue, 15 Sep 2015, Jason Gunthorpe wrote:
> >
> >> The mcast list in the core is soley for listing subscriptions for
> >> inbound - ie receive. Abusing it for send-side is probably the wrong
> >>
On Wed, Sep 16, 2015 at 10:59:50AM -0500, Christoph Lameter wrote:
> On Wed, 16 Sep 2015, Doug Ledford wrote:
>
> > > Abusing it for send-side is probably the wrong
> > > direction overall.
> >
> > I wouldn't "abuse" it for such, I would suggest adding a proper notion
> > of send-only
On 09/16/2015 11:59 AM, Christoph Lameter wrote:
> On Wed, 16 Sep 2015, Doug Ledford wrote:
>
>>> Abusing it for send-side is probably the wrong
>>> direction overall.
>>
>> I wouldn't "abuse" it for such, I would suggest adding a proper notion
>> of send-only registrations.
>
> That is really
RX checksum verification status is reported through wc_flag when polling CQ if
device supports checksum offload. When IBV_WC_IP_CSUM_OK is set, that means
both IPv4 header checksum and TCP/UDP checksum are OK.
TX checksum offload will be enabled for TCP/UDP over IPv4 if user sets
send_flag
Signed-off-by: Bodong Wang
---
src/mlx4.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/mlx4.c b/src/mlx4.c
index 2999150..9fe8c6a 100644
--- a/src/mlx4.c
+++ b/src/mlx4.c
@@ -203,8 +203,8 @@ static int mlx4_init_context(struct verbs_device
We are shifting by the _MASK macros instead of the _SHIFT ones.
Signed-off-by: Dan Carpenter
diff --git a/drivers/staging/rdma/hfi1/sdma.c b/drivers/staging/rdma/hfi1/sdma.c
index a8c903c..3a457d2 100644
--- a/drivers/staging/rdma/hfi1/sdma.c
+++
Another approach may be to tie the unsub from sendonly multicast joins to
the expiration of the layer 2 addresses in IPoIB. F.e. add code to
__ipoib_reap_ah() to detect if the handle was used for a sendonly
multicast join. If so unsubscribe from the MC group. This will result in
behavior
On Wed, 16 Sep 2015, Or Gerlitz wrote:
> On Wed, Sep 16, 2015 at 7:31 PM, Christoph Lameter wrote:
> > Another approach may be to tie the unsub from sendonly multicast joins to
> > the expiration of the layer 2 addresses in IPoIB. F.e. add code to
> > __ipoib_reap_ah() to detect
From: Ira Weiny
Some of the device files are required to be user accessible for PSM while
most should remain accessible only by root.
Add a parameter to hfi1_cdev_init which controls if the user should have access
to this device which places it in a different class with the
On Wed, Sep 16, 2015 at 11:17 PM, Christoph Lameter wrote:
> On Wed, 16 Sep 2015, Or Gerlitz wrote:
>
>> On Wed, Sep 16, 2015 at 7:31 PM, Christoph Lameter wrote:
>> > Another approach may be to tie the unsub from sendonly multicast joins to
>> > the expiration of
On Wed, Sep 16, 2015 at 7:31 PM, Christoph Lameter wrote:
> Another approach may be to tie the unsub from sendonly multicast joins to
> the expiration of the layer 2 addresses in IPoIB. F.e. add code to
> __ipoib_reap_ah() to detect if the handle was used for a sendonly
>
On Wed, 16 Sep 2015, Or Gerlitz wrote:
> Could you please post here a few (say 2-4) liner summary of what is
> still missing or done wrong in 4.3-rc1 and what is your suggestion how
> to resolve that.
With Doug's patch here the only thing that is left to be done is to
properly leave the
For RX: if corresponding QP is not supported, it will not validate the csum,
but packets are still received normally.
For TX: if corresponding QP is not supported for csum calculation and user
application sets the IBV_SEND_IP_CSUM flag, it will return error.
-Original Message-
From:
65 matches
Mail list logo